This project is designed to carry out specific research tasks relevant to and supportive of the overall, long- range goal of creating an RNA Ontology. The P.I. is spearheading the RNA Ontology Consortium (ROC), under the auspices of the RNA Society, which brings together interested scientists to develop an RNA Ontology that will integrate diverse data regarding RNA sequences, secondary and high-resolution 3D structures, and RNA function so as to create more usable knowledge about RNA for users in the biomedical community. This knowledge will take the form of fully integrated databases containing both sequence alignments and 3D structure and more powerful software to bring advanced computational methods to the bench scientist. It will facilitate and support the creation of highly specific and potent RNA-based therapeutics. A major component of the RNA Ontology will entail writing down definitions for all RNA motifs mentioned in the literature or appearing in databases, in a structured manner, agreed upon by the RNA community and interpretable by a computer. Thus, the specific aims of this project include: 1) To apply our new geometry-based RNA motif searching program, "Find RNA 3D" (FR3D), to search similar hairpin and internal loop motifs in the non-redundant database of high-resolution RNA 3D structures. 2) To use the discrepancy score calculated by FR3D for each pair of similar motifs, together with base-pairing annotations and other relevant features, to group motifs into families. 3) To identify common structural features within families that can serve as criteria for creating a systematic classification (taxonomy) of RNA motifs. This aspect of the work will be carried out in collaboration with other members of the ROC so as to achieve a consensus regarding classification issues. 4) To apply FR3D to identify new examples of non-Watson-Crick base-pairs in new 3D structures to update the Isostericity Matrices for each geometric base-pairing family. With FR3D we can also carry out context-specific searches. 5) To apply FR3D to identify and classify recurrent, isosteric base-stacking patterns in high-resolution crystal structures. To derive empirical base- stacking propensities for dinucleotides and to correlate these data with existing thermodynamic data. 6) To apply FR3D to identify structurally similar RNA tertiary interactions involving hairpin or internal loops interacting with minor groove receptors ("loop-receptor interactions"). [unreadable] [unreadable] [unreadable]